October 7th, 2013 – android, apps

Android, it’s time to take into account the dates on the app average rating!

Android was released 5 years ago, user rates from that time should weigh less than today’s rates on the app average rate.

I started coding for Android v1.5, something that few developers can say. I’ve seen the evolution all the way from that crappy apps with a horrendous design to the great apps that Android users currently enjoy. Google has also improved the way reviews are presented and now they even allow developers to answer to user reviews, something really useful to fix their problems.

However the app average rating is still calculated the same way. Just an average through all the ratings. No matter when the user posted the rating, no matter if the user also wrote a review, no matter if the user improved his rating over time, no matter if the developer has releases dozens of new versions. If a developer releases a bad first version, he is screwed unless he releases the ‘good’ fixed version under a new name.

Our app

On TouristEye, the company I co-founded, we released our iPhone travel guide app and our Android travel guide app at the same time in July 2010. We have evolved them a lot since then, 3 major versions to be specific (the last one with almost all fresh code). We got hundreds of thousands of downloads without spending a single cent on app marketing, and our users engage with the app for months. Our app of 2010 has nothing to do with our app of September 2013. But to calculate the average rating, Google takes into account ratings from all 3 years. Does it make sense?

We don’t have a bad average rating (3.98 with almost 2.900 ratings), but I’m pretty sure it doesn’t reflect the quality of the last version we launched. It’s just so amazing to read the last reviews, so many users we have helped recently. Just taking the last 40 reviews we have a 4.25 (and reviews are usually more incendiary than rates).

The solution

I don’t know how much the app average rating weighs on the position of the app in the Play Store, but it sure also has a psychological effect on users. I want that rating to be as real as possible.

So, if Google employees are so perfectionist, why don’t they fix this? Because yes, it’s a bug on their algorithm ;). Apple fixed it this summer, and we have noticed the changes for good as an important increment of downloads!

PS: I’d love to download a CSV with all the rates of my app to see the evolution.

There are really good comments in Hacker News.

Comments powered by Disqus